1123027319 P * stefani I'm Parting (the water)
1123028331 Q * tchan Ping timeout: 480 seconds
1123028659 J * tchan ~tchan@c-24-13-81-164.hsd1.il.comcast.net
1123029941 Q * tchan Ping timeout: 480 seconds
1123031870 J * tchan ~tchan@c-24-13-81-164.hsd1.il.comcast.net
1123033571 Q * Chand__ Quit: Computer goes to sleep!
1123035431 J * keyser_soze ~cimarron@host83.201-252-29.telecom.net.ar
1123035659 Q * neofutur Remote host closed the connection
1123036067 J * neofutur ~neofutur@neofutur.net
1123040996 N * pg|cereal cereal
1123041175 J * jpacheco ~justin@CPE0050180cd14b-CM0f0099806976.cpe.net.cable.rogers.com
1123041179 M * jpacheco they guys
1123041183 M * jpacheco oops
1123041185 M * jpacheco hey*
1123041459 M * jpacheco does anyone here know how i might get my guest to start its services on boot up?
1123041467 M * jpacheco im using gentoo as my guest os
1123042935 M * eyck that is known problems with gentoo guests
1123042941 M * eyck  fakeinit fixes that
1123042951 M * eyck I think it's documented somewhere
1123043210 M * jpacheco fakeinit?
1123043218 M * jpacheco is that an emerge?
1123043513 M * rhett just reading the guide saying that i should bind everything to a specific ip on the host machine
1123043529 M * rhett does that include portmap rpc.statd?
1123043535 M * rhett dhclient also
1123043536 M * rhett ?
1123044039 M * eyck jpacheco: it's configuration directive for vserver, you're using vserver, right?
1123044170 M * rhett can someone tell me if it is necessary for me to run port/rpc.statd i am trying to get everything to bind to specific ip addresses and unsure whether those ones are necessary?
1123044207 M * eyck rhett: you're talking about master vserver, or about guests maybe?
1123044247 M * rhett master vserver
1123044277 M * rhett i was reading a howto and it said on the master is hould bind everything to a specific ip problem is
1123044294 M * rhett tcp 0 0 *:679 *:* LISTEN 2193/rpc.statd
1123044294 M * rhett tcp 0 0 localhost.locald:sunrpc *:* LISTEN 1904/portmap
1123044307 M * rhett udp 0 0 *:673 *:* 2193/rpc.statd
1123044307 M * rhett udp 0 0 *:676 *:* 2193/rpc.statd
1123044307 M * rhett udp 0 0 *:bootpc *:* 1898/dhclient
1123044318 M * rhett these ones are on *: not a specific ip so wondering if i need them
1123044332 J * Matthew-1 ~mlong@grm-nx.grmims.com
1123044378 M * jpacheco eyck: yeah, im using vserver
1123044389 M * jpacheco eyck: so what setting do i have to change?
1123044394 M * Matthew-1 hello
1123044405 M * jpacheco eyck: or where can i find the documentation on this?
1123044426 M * Matthew-1 i have an issue
1123044442 M * Matthew-1 is any one there
1123044489 M * Matthew-1 has any one gotten vserver to work on sles9
1123044581 M * Matthew-1 i ran the testme.sh scripted from the 2.6 kernel vserver howto
1123044601 M * Matthew-1 and three tests failed.
1123044625 M * Matthew-1 [000]# failed.
1123044635 M * Matthew-1 [031]# failed.
1123044641 M * Matthew-1 [202]# failed.
1123044661 M * Matthew-1 any help would be appreciated 
1123044665 M * Matthew-1 thank you
1123044824 M * keyser_soze the -v parameter will show you wich tests are failing
1123044828 M * Matthew-1 this is on an opteron box x86_64
1123044834 M * Matthew-1 ok
1123044839 M * Matthew-1 i'll try that
1123044874 M * Matthew-1 i looked at the scripted all ready
1123044894 M * Matthew-1 chcontext true && chcontext --ctx 45678 true
1123044904 M * Matthew-1 that failed
1123044918 M * Matthew-1 i think i need to read more
1123044930 M * Matthew-1 i don't know what that means
1123045361 Q * keyser_soze Quit: Abandonando
1123045377 Q * Matthew-1 Quit: Matthew-1
1123045517 J * keyser_soze ~cimarron@host83.201-252-29.telecom.net.ar
1123045709 M * eyck jpacheco: http://www.paul.sladen.org/vserver/archives/200211/0059.html
1123045729 M * eyck jpacheco: "How To Get A Gentoo Vserver To Work"
1123045846 Q * _ag_ Ping timeout: 480 seconds
1123046071 Q * keyser_soze Quit: Abandonando
1123046243 J * keyser_soze ~cimarron@host83.201-252-29.telecom.net.ar
1123048605 M * eyck rhett: that was just geeneral rule, it is possible that you don't need those services on master at all, but dhclient, that you probably need
1123048620 M * eyck and you won't be using it inside guests so there's no problem here.
1123050244 Q * maharaja Ping timeout: 480 seconds
1123050247 N * Bertl_zZ Bertl
1123050267 M * Bertl morning folks!
1123050616 M * keyser_soze good morning 
1123050972 M * rhett hmn
1123051013 M * Bertl rhett: regarding the binding: the thing is basically this ...
1123051045 M * Bertl if you use (directly or indirectly) the chbind command, it will restrict bindings to a given set of ip addresses
1123051078 M * Bertl by default, the host (root/master) context does not use chbind, but usually all guests(vservers) use it ...
1123051106 M * Bertl so if a service on the host binds to 0.0.0.0 port 42, then no guest will be able to bind to that same port
1123051138 M * rhett if i setup one of my virtual servers and i just make straight copies of the directory can i launch that as another virtual server?
1123051142 M * Bertl specifying the host ip only, or using chbind for services on the host (with the v_* wrappers) will remove that limitation
1123051167 M * rhett so if i setup one server with all the stuff i want i can just copy that to another directory name and launch a new instance of that?
1123051170 M * Bertl rhett: with a proper config yes
1123051179 M * rhett cool
1123051189 M * Bertl it's called template (the prepared guest which isn't used)
1123051214 M * rhett so i login to the template one install all the stuff
1123051218 M * rhett then copy it and the config for it?
1123051235 M * Bertl so with the new tools you basically to it like this (until the copy build method is done)
1123051256 M * Bertl vserver <new-name> build -m skeleton ... <options> ...
1123051257 M * rhett i installed the newest dev version
1123051268 M * Bertl (this will create an empty but configured guest)
1123051273 M * rhett k
1123051283 M * rhett and skeleton is just the name of the vserver i want to use as a template?
1123051285 M * Bertl then copy over the template to the root dir (or just replace it)
1123051309 M * Bertl no, skeleton is the build method, <new-name> is the name of the new guest
1123051320 M * rhett ah k cool
1123051331 M * Bertl as I said, the template copy is not done yet in the tools but it's fairly trivial
1123051364 M * rhett i installed the newest kernel vtools thign seems to all be working fine so far
1123051443 M * Hollow morning
1123051479 M * Aiken Bertl I noticed your name on the qemu mailing list
1123051495 M * Aiken have you played much with qemu + sparc?
1123051501 M * Bertl Hollow: morning!
1123051520 M * Bertl Aiken: not yet .. but I'm following the sparc development for some time
1123051535 M * Bertl Aiken: and I'm using QEMU for development ...
1123051539 M * Aiken I was hoping you have done a lot and have a working bootable image
1123051564 M * Bertl you are talking of sparc as guest or host?
1123051569 M * Aiken guest
1123051600 M * Bertl thought so ... well, we can give it a try if you like ... I'm interested to spend some time on that ...
1123051702 M * Aiken http://lists.gnu.org/archive/html/qemu-devel/2004-02/msg00162.html
1123051713 M * Aiken that is an example of what I have done with qemu + ppc
1123051726 M * Aiken but have never managed simillar with sparc and sparc is what I want
1123051800 M * Bertl k, but that is binary emulation, not full system
1123051821 M * Aiken I know
1123051850 M * Aiken but if I can not even get that far with sparc what is the use
1123051863 M * Aiken plus with full system emulation I have not managed to get a kernel to boot yet
1123051876 M * Aiken which is why I was wondered how far you have gotten
1123051923 M * rhett can someone paste me their interfaces file for vserver pls
1123052067 M * Bertl rhett: hmm, what do you want?
1123052230 J * erwan_ho ~erwan@konilope.dyndns.org
1123052271 M * rhett hmn
1123052286 M * rhett i need to setup an interface with the gateway etc on one of the vservers
1123052307 M * Bertl you do not do networking on the guests ...
1123052313 M * rhett i was looking at some page with marijuana leaves all over it that some guide refers to hehe but bit confusing
1123052319 M * rhett hmn
1123052323 M * Bertl basically all net working is done on the host ...
1123052324 M * rhett how do i allow a guest to get access to an ip?
1123052363 M * Bertl so you setup everything properly on the host, then either assign the ip to the guest, or allow the tools to create the ip/alias for you
1123052378 M * Bertl (the second one will automatically remove the ip/alias on shutdown)
1123052400 M * rhett whats the tool to assign an ip?
1123052415 M * Bertl Aiken: nothing yet, as I said, but I think the only issue will be the openprom ...
1123052437 M * rhett eek i need to catch ferry will bbl thanks for all help
1123052442 M * Bertl rhett: if you build a config with the beforementioned skeleton config
1123052451 M * Bertl rhett: okay, later :)
1123052694 M * Bertl Aiken: let me get some recent qemu version ... and look for an openprom image ...
1123053267 M * Aiken 1 out of 5 tries it tried booting a kernel
1123053705 M * Bertl okay, building 0.7.1 for all sys archs ...
1123053967 M * meebey morning!
1123054009 M * meebey Bertl: I will go now the next stage
1123054033 M * meebey Bertl: setting the network/bcast stuff with new config structure
1123054041 M * Bertl good ...
1123054064 M * meebey because the boot stage is essential
1123054081 M * meebey samba registers at the WINS server during that, and that fails
1123054369 Q * erwan_ho Remote host closed the connection
1123054711 Q * keyser_soze Quit: Abandonando
1123054785 J * keyser_soze ~cimarron@host83.201-252-29.telecom.net.ar
1123055201 M * Bertl hmm .. okay after some startup issues (compiling 0.7.1) now I've got all archs (emuls for qemu) now let's see how x86 works (as test)
1123055249 M * xf_ postfix: fatal: could not find any loopback addresses
1123055249 M * xf_ yo, any idea how to fix that (what i pasted above) please?
1123055278 M * Aiken if you ignore the erratic clock, x86 works fine
1123055286 M * Bertl xf_: hmm, care to share some details about host/guest/setup?
1123055306 M * xf_ debian host, ubuntu guest
1123055316 M * xf_ um, pretty stock standard postfix config, really
1123055327 M * Bertl xf_: okay, please give the testme.sh a spin, you know where to get it?
1123055335 M * xf_ is there a way to get a loopback interface up, given it's all part of the vserver magic?
1123055338 M * xf_ no idea
1123055369 M * Bertl http://vserver.13thfloor.at/Stuff/SCRIPT/testme.sh
1123055386 M * Bertl Aiken: well, for one, the qemu-fast does not work with 0.7.1 :/
1123055418 M * Bertl but the softmmu one does ...
1123055464 M * xf_ ok, i'm running it, what should i expect?
1123055480 M * Bertl some kind of output, you please upload somewhere (e.g. pastebin.com)
1123055496 M * xf_ how long does it take?
1123055498 M * xf_ it's taking a while
1123055509 M * Bertl no, not really, should be done in a few seconds
1123055514 M * xf_ root@bluepanel:~# ./testme.sh -v
1123055515 M * xf_ Linux-VServer Test [V0.13] Copyright (C) 2003-2005 H.Poetzl
1123055517 M * xf_ just sitting there
1123055540 M * xf_ grep -q TELL_UNSAFE_CHROOT `which chcontext`
1123055540 M * xf_ which chcontext
1123055542 M * xf_ it's hanging there
1123055550 M * Bertl well, then something is wrong, no?
1123055560 M * xf_ beats me what/how
1123055574 M * Bertl probably chcontext is not in the path
1123055583 M * Bertl are you executing it as root on the host?
1123055596 M * xf_ oh, right, no
1123055599 M * xf_ go me :P
1123055616 M * Bertl but a good test, I'll add some checks there in the next release
1123055676 M * xf_ http://pastebin.com/328190
1123055742 M * Bertl hmm, interesting ...
1123055759 M * Bertl xf_: okay, when you have started the guest, can you enter it?
1123055790 M * Bertl Aiken: do you have some proll.bin?
1123055810 M * Aiken no, I am using the -kernel option
1123055824 M * Bertl yes, me too, but it requires the openprom ...
1123055847 M * Aiken oh, I wonder I had 1 kernel start
1123055921 M * Aiken         ^ why
1123055926 M * xf_ Bertl: sure
1123055940 M * xf_ just, postfix is choking because it can't find lo, i'm guessing
1123055944 M * Bertl Aiken: no idea, I get: qemu) qemu: could not load prom '/usr/local/share/qemu/proll.bin
1123055990 M * Bertl xf_: okay, then please enter it, and try 'cat /proc/self/ninfo'
1123056016 M * Hollow Bertl: you'd be available on friday?
1123056016 M * Aiken using strace I notice is looking for and finding proll.elf
1123056020 M * Aiken no error message
1123056024 M * xf_ Bertl: ok, done that, what am i looking for?
1123056034 M * Bertl Hollow: sure, in Vienna?
1123056038 M * Hollow yup
1123056042 M * Bertl xf_: again, the output :)
1123056053 M * xf_ ok, i'm going to change the actual ip addr though for obvious reasons
1123056063 M * Bertl Hollow: yeah, should be no problem ...
1123056073 M * Bertl xf_: that's fine, just use xxx.xxx ...
1123056077 M * xf_ (not because of you, just because pastebin is public :p)
1123056084 M * Hollow there is some restaurant in vienna which sells schnitzel as big as a pizza, you know it?
1123056098 M * Bertl yep, I do .. Schnitzelwirt :)
1123056108 M * Hollow well, i'd like to have one ;) it's been a long time.. ;)
1123056120 M * xf_ http://pastebin.com/328192
1123056138 M * xf_ btw, that side of things works fine - i've got sshd installed on the guest, and i'm logged into it via ssh
1123056139 M * Bertl Hollow: k, just call me when you are in range (see query)
1123056394 M * xf_ any idea?
1123056418 M * xf_ does postfix require any special config for vserver, or should i be getting some kind of loopback interface?
1123056429 M * xf_ won't other daemons choke because they can't find lo?
1123056492 M * Bertl as far as I know postfix works fine without any changes
1123056518 M * Bertl but it might be that the ubuntu/debian people do some 'special' setup, which in turn requires a loopback interface
1123056574 M * Bertl xf_: best would be to check the postfix config for references to 127.0.0.1 or lo
1123056635 M * meebey Bertl: V-Server 2.0 will be kernel 2.6 only right?
1123056644 M * Bertl yep, precisely
1123056656 M * meebey hm so I have to update to 2.6 soon
1123056711 M * Bertl well, 1.2.x is also stable ... so if you don#t need the features of 2.0 .. you can sty with 2.4
1123056736 M * meebey I plan to migrate to 2.6 anyhow, so
1123056846 M * meebey Bertl: has 2.0 a feature that allows proc write access?
1123056881 M * Bertl no, not yet ...
1123056887 M * meebey ic
1123056900 M * meebey so I need to port my patch to 2.0 then
1123056942 M * Bertl well, I would check it first, 2.0/2.6 is pretty different in many ways ...
1123056963 M * Bertl Aiken: yay! http://www.hadrons.org/~guillem/debian/pool/main/proll/
1123057009 J * Doener ~doener@p548776E2.dip.t-dialin.net
1123057015 M * Bertl morning Doener!
1123057022 M * Doener morning Bertl
1123057032 M * meebey Bertl: ic
1123057054 M * meebey Bertl: the docs on linux-vserver.org should be seperated more IMHO
1123057056 M * meebey Bertl: into vs1.2.x and vs1.9/2.0 sections
1123057069 M * meebey new users are confused what doc is for what version
1123057069 A * Doener .oO( it's a wiki )
1123057076 M * Doener ;)
1123057115 M * Bertl meebey: go crazy (if you can change it for the better) but make use of the preview option
1123057167 M * Bertl Aiken: now I get a pretty nice register/core dump
1123057182 M * Aiken I have not down load that file yet
1123057227 M * xf_ Bertl: i have, no references.
1123057229 M * Aiken I am now gettting this far http://pastebin.com/328198
1123057270 M * Bertl hey cool, how so? I was using the sparc image provided by fabrice ...
1123057312 M * Aiken latest qemu from cvs
1123057325 M * Aiken compiled a 2.6.12.3
1123057346 M * Aiken it comes up as a sun4m
1123057673 M * xf_ hmm is uml going to give me these kind of issuesi
1123057684 M * xf_ i'm noticing it's not really like a true virtual server, i'm having to hack around all sorts of things
1123057689 M * meebey galilei:/etc/vservers# cat smb/interfaces/dev 
1123057689 M * meebey eth0
1123057702 M * xf_ traffic graphing for example isn't accurately working, it's counting the host,not just the guests
1123057703 M * meebey that is not enough to have the IPs too of that interface in the vserver?
1123057721 M * meebey I want something like I had with IPROOT=eth0 before
1123057780 M * meebey "ip addr" shows no IPs of that interface
1123057850 M * Bertl the 'ips' of that interface does not make much sense actually
1123057865 M * meebey no?
1123057872 M * Bertl because the interface might be changed any time ... and the guest can not follow those settings
1123057891 M * meebey right, but having thet IPs that was there during vsever startup is ok
1123057895 M * meebey thats how it worked bfore
1123057902 M * Bertl I don't know if the tools support something like that (for the new style) but it's basically deprecated
1123057911 M * meebey hmpf
1123057931 M * Bertl if you are happy with the ips there when you startup, why not just write them to the config?
1123057935 M * meebey seems like I am stuck between old deprecated and new less featured tools
1123057954 M * Bertl and of course, you can always file a feature request to savannah
1123057957 M * meebey Bertl: the IPs change often, like PPP reconnect
1123057972 M * Bertl well, in this case the guest will magically fail, no?
1123057972 M * meebey Bertl: also I don't know the IPs, they are general v-server system, the IP setup is part of the host
1123058155 M * Bertl well, the usual approach would be to run a simple script (on startup or whenever the host ips change) and do whatever necessary (change the guest setup, restart guests)
1123058189 M * Bertl or, as a much simpler approach, run the guests on private ips, and do the network access via SNAT ...
1123058203 M * Bertl (doesn't require to restart any guest when the host ip changes)
1123058474 M * Bertl xf_: that's expected ... because you _are_ graphing the host
1123058500 M * Bertl xf_: if you want to graph guest traffic, you have to do some iptables accounting rules for that ...
1123058531 M * Bertl (or wait for ngnet, which will have that and much more, e.g. virtual lo devices)
1123059167 M * Bertl Aiken: where did you get your proll.bin from?
1123059209 M * Aiken I don;t have a proll.bin
1123059229 M * Aiken that is where I am upto http://pastebin.com/328212
1123059233 M * Bertl hum ...
1123059250 M * Aiken (root@fred) find /opt/apps/qemu/ -name 'proll*'
1123059250 M * Aiken /opt/apps/qemu/share/qemu/proll.elf
1123059272 M * Bertl where did that come from?
1123059314 M * Aiken qemu sources, in the pc-bios directory
1123059342 M * Bertl indeed, it's there ...
1123059380 M * Bertl ah, and it works much better ...
1123059493 J * mastachand222 ~chand@m157.net81-64-156.noos.fr
1123059589 M * Bertl Aiken: it seems that the disk emulation is pretty broken
1123059614 M * Aiken I had to enable scsi
1123059625 M * Aiken Sparc ESP Scsi Driver
1123059643 M * Bertl I'm still testing the provided kernel
1123059652 M * Bertl (sparc-test) and it sees my disk images
1123059663 M * Bertl but access to a partition (etx2) results in 
1123059667 M * Bertl sda1: rw=0, want=146933154, limit=65457
1123059667 M * Bertl attempt to access beyond end of device
1123059681 M * Bertl (but the image is quite fine ... it works on x86)
1123059705 M * Aiken at the moment I am getting        *** glibc detected *** malloc(): memory corruption: 0x09bc8740 ***
1123062339 M * Hollow lalala
1123062352 M * Bertl hmm?
1123062367 M * Hollow http://phpfi.com/72498 :)
1123062368 M * eyck bum bum
1123062377 M * Hollow just completed automatic stage3 vserver installation ;)
1123062415 M * Doener nice :)
1123062439 M * Hollow http://home.xnull.de/work/gentoo/vserver/tools/vserver-new
1123062501 M * Hollow boots out-of-the box with baselayout-vserver and plain initstyle :)
1123062578 M * Bertl sounds good ...
1123062614 M * Hollow and only took about 5 minutes on my new box :)
1123062620 M * Hollow to build
1123064029 M * meebey seems like I can cheat a bit
1123064041 M * meebey smb/interfaces/0/dev: eth0
1123064054 M * meebey smb/interfaces/0/ip: 192.168.8.111
1123064059 M * meebey smb/interfaces/0/nodev
1123064087 M * meebey the old way, passing 1:1 interfaces and with that all IPs of it, I like better
1123064098 M * meebey now I have to name the IPs
1123064110 M * meebey s/name/specify/
1123064380 J * mef ~mef@pcp09895218pcs.ewndsr01.nj.comcast.net
1123064387 M * mef hi bertl
1123064396 M * Bertl hey mef! how are you?
1123064415 M * mef Am in the process of upgrading from 1.9.5.x5 to 2.0-pre1 and those that follow.
1123064419 M * mef I am fine.
1123064429 M * mef just have a question...
1123064434 M * Bertl ah, well, rc9 it the current ...
1123064457 M * mef I know, but I like to go through each stage to see what has changed.
1123064469 M * mef until I get to rc9.
1123064482 M * Bertl we have a changelog :)
1123064488 M * mef doah.. 
1123064525 M * Bertl http://linux-vserver.org/ChangeLog26
1123064535 M * Bertl just the last one is missing .. yet
1123064555 M * mef looking at it
1123064633 M * mef ok
1123064643 M * Aiken Bertl while I think of it, the alpha with rc9 has been up 2 days and 8 min
1123064655 M * Bertl Aiken: excellent!
1123064665 M * Aiken spent 8 - 10 hours with both guest and host compiling 
1123064682 M * Aiken enough activity the system load stayed > 2
1123064690 M * meebey Bertl: I found my broadcast problem, faulty firewall that was
1123064697 M * jpacheco hey guys
1123064712 M * meebey Bertl: sorry for the annoyance
1123064725 M * Bertl meebey: good to hear ...
1123064736 M * jpacheco rc9 is pretty good for me so far
1123064738 M * Bertl hey jpacheco!
1123064750 M * jpacheco hey Bertl!
1123064759 M * Bertl meebey: about 98% of all 'vserver' issues are not vserver related ...
1123064852 M * meebey Bertl: :) servers can be complex and vserver adds another layer of that :)
1123064868 M * Bertl definitely ...
1123064888 M * Bertl and of course if folks try them out, they naturally look for issues there ...
1123064897 M * meebey yep
1123064907 M * meebey poor Bertl 
1123064918 M * jpacheco hey Bertel
1123064927 M * meebey but thanks for good support and help, keep it like that! :)
1123064933 M * Bertl ah, well, I learn a lot with those issues ...
1123064934 M * jpacheco are you aware that vserver-copy is looking or $vseverName.conf
1123064945 M * Bertl meebey: you're welcome, and I will ...
1123064956 M * Bertl jpacheco: what is vserver-copy? *G*
1123064961 M * jpacheco lol
1123064973 M * jpacheco its a script that is packaged with util-vserver
1123064980 M * Bertl really?
1123064984 M * jpacheco it allows you to copy two vservers
1123064986 M * jpacheco yeah
1123064994 M * jpacheco well i got it with gentoo
1123065001 M * jpacheco when i emerged util-vserver
1123065003 M * meebey Bertl: where can I requests features? I want anoynmous interface support to have all IPs of it in the vserver without specifying them :-P
1123065038 M * Bertl jpacheco: indeed, found it in the legacy scripts ...
1123065059 M * Bertl should probably be removed soon (including the legacy tools)
1123065139 M * jpacheco well i was wondering if i could replace it
1123065143 M * meebey Bertl: please don't :)
1123065162 M * meebey or it needs some more features
1123065181 M * Bertl jpacheco: if you want to 'improve' this, add some real tool support for a copy target
1123065189 M * meebey I would like to the use the new tools but currently they can't do the work that the old tools did
1123065203 M * Bertl of course, that might include support for legacy configs
1123065212 M * jpacheco real tool support?
1123065228 M * Bertl you know util-vserver 'vserver' tool?
1123065241 M * Bertl vserver <name> build and such?
1123065268 M * jpacheco yeah
1123065287 M * meebey Bertl: pre-start does not work like post-stop
1123065299 M * Bertl jpacheco: check out 'vserver - build --help'
1123065314 M * jpacheco i've done that many times
1123065315 M * jpacheco hahahah
1123065316 M * meebey Bertl: calling "vserver X running" in pre-start will say it's not running
1123065329 M * meebey Bertl: post-stop tells its running, postpost-stop not
1123065341 M * meebey Bertl: so pre-start is actually prepre-start
1123065353 M * Bertl meebey: could you file that to savannah?
1123065362 M * meebey sure
1123065365 M * Bertl tx
1123065383 M * Bertl jpacheco: okay, so some 'clone' build method there would be really nice ...
1123065384 M * jpacheco did u want vserver copy to be writen into vserver - build or something?
1123065390 M * jpacheco ahhh
1123065404 M * Bertl especially as the feature is missing right now ...
1123065426 M * jpacheco do you have a list of features that you would want met with this addition?
1123065495 M * Bertl it would be nice if the 'script' (and it's nothing more than a script) would allow to modify the settings/attributes of the copy (by passing certain config options, as for the build)
1123065515 M * Bertl IIRC, I even did some stub helper for this and/or the guest removal ...
1123065563 M * Bertl let me see if I can find them ...
1123065613 M * Bertl http://vserver.13thfloor.at/Experimental/delta-vserver-bertl01.diff
1123065628 M * jpacheco brb, gotta go to work
1123065631 M * jpacheco i'll be in later tonight
1123065633 M * jpacheco later
1123065634 M * Bertl k, cya!
1123065652 Q * jpacheco Quit: BitchX-1.1-final -- just do it.
1123067152 Q * albeiro Ping timeout: 480 seconds
1123067208 M * Bertl okay, I'm off for now .. will be back later ...
1123067208 J * albeiro ~albeiro@graffias.estrefa.pl
1123067225 N * Bertl Bertl_oO
1123068770 M * rhett back
1123068775 M * rhett oh know i missed bertl:p
1123068781 M * rhett oh no even heh
1123068813 Q * Aiken Quit: Leaving
1123068958 M * rhett how do i allow my vserver to use an ip address?
1123068988 M * TheSeer ??
1123068994 M * TheSeer define "use"
1123069010 M * rhett like if i go vserver box enter
1123069018 M * rhett then ifconfig or ifup i cant get to any network interface
1123069024 M * TheSeer ah
1123069027 M * rhett how to allow from within a vserver to have an interface
1123069029 M * TheSeer no, you have to do that from outside
1123069047 M * TheSeer allowing that INSIDE a veserver would be a security nightmare
1123069051 M * rhett hmn
1123069061 M * rhett how do i allow the vserver to have internet or bind services to an ip address?
1123069073 M * rhett at the moment it seems that internet is not working from within this vserver?
1123069109 M * TheSeer what ip did you give your vserver guest?
1123069127 M * rhett i didnt i dont think
1123069140 M * rhett vserver box build -m debootstrap -- -d sid -m ftp://ftp.au.debian.org/debian/
1123069144 M * rhett i did that to make it
1123069149 M * rhett then vserver box enter
1123069154 M * TheSeer okay..
1123069162 M * TheSeer amazing it actually worked ;>
1123069166 M * rhett heh
1123069184 M * rhett what should i have done ?:>
1123069190 M * TheSeer you should have had to at least run vserver box start before you could enter it
1123069200 M * rhett oh i did that also sorry hehe
1123069207 M * rhett vserver box start ; vserver box enter
1123069207 M * TheSeer go to /etc/vservers
1123069222 M * rhett k
1123069224 M * TheSeer there's a directory structure for each vserver you create
1123069255 M * TheSeer it should have an "interfaces" folder
1123069265 M * rhett found that
1123069286 M * TheSeer okay.. you got 2 files in there - or should have 
1123069295 M * TheSeer 1.  dev 
1123069314 M * TheSeer it shold contain the interface device, like eth0 eth1 or dummy0 or whatever you use
1123069318 M * TheSeer 2. mask
1123069329 M * rhett hmn its empty that folder sec
1123069339 M * TheSeer k.. then create those files
1123069345 M * TheSeer echo "eth0" > dev
1123069353 M * TheSeer echo "255.255.255.0" > mask
1123069361 M * TheSeer or whatever the real hostmask would be
1123069370 M * TheSeer then create a new subfolder
1123069377 M * TheSeer 0  for the 1st ip
1123069378 M * rhett so i can do echo "192.168.1.1" > gateway
1123069379 M * rhett ?
1123069397 M * TheSeer ouhm.. not sure if that would make any sense
1123069404 M * rhett k
1123069406 M * rhett mkdir 0
1123069407 M * rhett what now heh
1123069414 M * TheSeer echo "1.2.3.4" > 0/ip
1123069426 M * TheSeer wheres 1.2.3.4 would be the real ip you want your vserver to use
1123069446 M * rhett do i set gateway here?
1123069458 M * rhett or it figures gateway from base setting?
1123069509 M * TheSeer technically, the vserver "interface" is an alias to eth0
1123069517 M * TheSeer so it will use the "host" routing
1123069534 M * rhett k
1123069547 M * TheSeer dunno if bertl changed that for the tng networking though
1123069563 M * rhett nice it works
1123069568 M * rhett thanks ;)
1123069595 M * TheSeer you're welcome :)  you could have added the parameters for the interface and ip to the build call btw
1123069633 M * rhett ah k
1123069654 M * rhett i was a bit dubious i came on the other night and someone pointed me to the alpha of everthing but it all seems to be working hehe
1123069665 M * rhett when i am in the host system and i do shutdown -r now
1123069666 M * rhett etc
1123069676 M * rhett will that safely shutdown the vservers or i need to put some shutdown scripts in there?
1123069715 M * TheSeer i'm not sure it'll do anything at all ;)
1123069737 M * TheSeer technically, a vserver is nothing but a seperate context within the main kernel
1123069761 M * TheSeer so if the last process within that context did end, the "vserver" is pretty much gone too
1123069781 M * TheSeer (yes, bertl, that's not 100% accurate ;-P )
1123069838 Q * mef Quit: using sirc version 2.211+KSIRC/1.3.10
1123070987 M * rhett postgresql: Depends: libkrb53 (>= 1.3.2) but it is not going to be installed
1123070987 M * rhett Depends: libperl5.8 (>= 5.8.4) but it is not going to be installed
1123070998 M * rhett when trying to use apt-get it doesnt want to install any of the dependencies
1123071005 M * rhett is this a permission thign i can set somewhere?
1123071555 M * meebey sounds like testing/unstable, your problem :)
1123071800 M * rhett whats that?:>
1123071840 M * meebey debian testing/unstable
1123071902 M * rhett what do i do to fix that ?
1123072207 M * rhett help!:)
1123072274 M * rhett how to tell it to ignore if somethign is marked as unstable?
1123072300 M * rhett it doesnt do this to me on the host machine?
1123072324 M * meebey is that debian or not?
1123072330 M * rhett its debian sarge
1123072383 M * rhett i think lol
1123072410 M * rhett maybe i messed that up lol i go check
1123072434 M * rhett sid is another debian daemon?
1123072582 M * meebey " but it is not going to be installed" means the dependencies can't be fulfilled
1123072589 M * meebey that does not happen with debian/stable
1123072591 M * meebey aka sarge
1123072602 M * meebey it has nothing todo with vserver though
1123072618 M * meebey so ask on some debian channel instead
1123072654 M * rhett ah k
1123072904 M * rhett yeah i think i made stupid mistake trying to be fast copy/pasting from teh howto guide
1123072909 M * rhett think i had the wrong distribution on there lol
1123073447 Q * keyser_soze Quit: Abandonando
1123075943 M * eyck w
1123075949 J * Secti0n8 ~ody@host81-156-184-240.range81-156.btcentralplus.com
1123076462 J * keyser_soze ~cimarron@host83.201-252-29.telecom.net.ar
1123076925 P * Secti0n8 
1123077312 J * mef ~mef@targe.CS.Princeton.EDU
1123078377 Q * albeiro Ping timeout: 480 seconds
1123079452 J * albeiro ~albeiro@graffias.estrefa.pl
1123084650 M * sid3windr anyone know if asterisk runs under vserver?
1123085590 Q * mastachand222 Quit: Quitting!
1123086178 N * Bertl_oO Bertl
1123086189 M * Bertl evening folks!
1123086203 M * Bertl sid3windr: yes, some folks actually use it ... but with certain caps IIRC
1123086369 M * sid3windr nice
1123086574 M * Bertl mef: I added the rc9 to the changelog too ...
1123086722 J * lilo_ ~lilo@lilo.usercloak.oftc.net
1123086736 M * Bertl wb lilo_! fading recently?
1123086844 Q * lilo Ping timeout: 480 seconds
1123087680 Q * keyser_soze jupiter.oftc.net arion.oftc.net
1123087680 Q * Doener jupiter.oftc.net arion.oftc.net
1123087680 Q * tchan jupiter.oftc.net arion.oftc.net
1123087680 Q * duckx jupiter.oftc.net arion.oftc.net
1123087680 Q * DaPhreak jupiter.oftc.net arion.oftc.net
1123087680 Q * SNy jupiter.oftc.net arion.oftc.net
1123087680 Q * micah jupiter.oftc.net arion.oftc.net
1123087680 Q * mugwump jupiter.oftc.net arion.oftc.net
1123087680 Q * prae jupiter.oftc.net arion.oftc.net
1123087680 Q * stupidawy jupiter.oftc.net arion.oftc.net
1123087680 Q * revenger_ jupiter.oftc.net arion.oftc.net
1123087680 Q * neofutur jupiter.oftc.net arion.oftc.net
1123087680 Q * monrad jupiter.oftc.net arion.oftc.net
1123087680 Q * sid3windr jupiter.oftc.net arion.oftc.net
1123087680 Q * FaUl jupiter.oftc.net arion.oftc.net
1123087680 Q * meebey jupiter.oftc.net arion.oftc.net
1123087680 Q * derbien jupiter.oftc.net arion.oftc.net
1123087680 Q * flock jupiter.oftc.net arion.oftc.net
1123087680 Q * chand jupiter.oftc.net arion.oftc.net
1123087680 Q * virtuoso jupiter.oftc.net arion.oftc.net
1123087680 Q * sladen jupiter.oftc.net arion.oftc.net
1123087680 Q * BobR_oO jupiter.oftc.net arion.oftc.net
1123087680 Q * skceb jupiter.oftc.net arion.oftc.net
1123087680 Q * eyck jupiter.oftc.net arion.oftc.net
1123087680 Q * anonymousc jupiter.oftc.net arion.oftc.net
1123087680 Q * TheSeer jupiter.oftc.net arion.oftc.net
1123087680 Q * Hollow jupiter.oftc.net arion.oftc.net
1123087680 Q * ray6 jupiter.oftc.net arion.oftc.net
1123087680 Q * cryo jupiter.oftc.net arion.oftc.net
1123087680 Q * Pazzo jupiter.oftc.net arion.oftc.net
1123087680 Q * rhett jupiter.oftc.net arion.oftc.net
1123087680 Q * Bertl jupiter.oftc.net arion.oftc.net
1123087680 Q * dsoul jupiter.oftc.net arion.oftc.net
1123087680 Q * mcp jupiter.oftc.net arion.oftc.net
1123087680 Q * locksy jupiter.oftc.net arion.oftc.net
1123087680 Q * janra jupiter.oftc.net arion.oftc.net
1123087710 T * services.oftc.net http://linux-vserver.org/ | latest stable 1.2.10, 2.0-rc9, ngnet ng9.5  -- He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;)
1123087717 J * neofutur ~neofutur@neofutur.net
1123087717 J * monrad ~monrad@213083190130.sonofon.dk
1123087717 J * rhett ~lala@dsl-58-6-34-71.nsw.westnet.com.au
1123087717 J * Pazzo ~Pazzo@host130-250.pool8172.interbusiness.it
1123087717 J * cryo ~say@212.86.243.154
1123087717 J * ray6 ~ray@klon3.gcsc2.ray.net
1123087717 J * Hollow ~Hollow@home.xnull.de
1123087717 J * mcp ~hightower@wolk-project.de
1123087717 J * TheSeer ~theseer@border.office.salesemotion.net
1123087717 J * anonymousc ~anonymous@staff.internode.com.au
1123087717 J * eyck eyck@81.219.64.71
1123087717 J * sid3windr luser@bastard-operator.from-hell.be
1123087717 J * FaUl ~immo@hobbynuttenverzeichniss.de
1123087717 J * dsoul darksoul@vice.ii.uj.edu.pl
1123087717 J * meebey meebey@booster.qnetp.net
1123087717 J * derbien ~derbien@whiterabbit.nbmc.de
1123087717 J * locksy ~locksy@mrtg.sisgroup.com.au
1123087717 J * flock ~restless@l192-117-111-12.broadband.actcom.net.il
1123087717 J * chand ~chand@staff.lycos.fr
1123087717 J * virtuoso ~s0t0na@80.253.205.251
1123087717 J * sladen paul@starsky.19inch.net
1123087717 J * BobR_oO ~georg@212.16.62.52
1123087717 J * Bertl ~herbert@212.16.62.52
1123087717 J * skceb skceb@bang-bang.feuer-frei.com
1123087717 J * janra janra@paradox.homeip.net
1123087717 T * xenon.oftc.net http://linux-vserver.org/ | latest stable 1.2.10, 2.0-rc9, ngnet ng9.5  -- He who asks a question is a fool for a minute; he who doesn't ask is a fool for a lifetime -- share the gained knowledge on the wiki, and we'll forget about the minute ;)
1123087737 J * keyser_soze ~cimarron@host83.201-252-29.telecom.net.ar
1123087737 J * Doener ~doener@p548776E2.dip.t-dialin.net
1123087737 J * tchan ~tchan@c-24-13-81-164.hsd1.il.comcast.net
1123087737 J * revenger_ ~joe@bulldog.infosys.de
1123087737 J * stupidawy foo@198.77.239.131
1123087737 J * prae ~prae@gut75-1-81-57-27-189.fbx.proxad.net
1123087737 J * mugwump ~samv@210-54-92-184.ipnets.xtra.co.nz
1123087737 J * duckx ~Duck@mna75-1-81-57-39-234.fbx.proxad.net
1123087737 J * DaPhreak ~phreak@styx.xnull.de
1123087737 J * SNy a7c816ce0c@bmx-chemnitz.de
1123087737 J * micah micah@micha.hampshire.edu
1123088060 M * Bertl everbody back in one piece?
1123088077 Q * prae Quit: Execute Order 69 !
1123088536 M * brc BERTLLLL
1123088548 M * Bertl BRCCC!
1123088584 M * brc sorry for bothering you with this. i had to give cap_sys_resource to a user. would that compromise the host machine secuirty ?
1123088604 M * Bertl have you read the comments in the capabilities.h file?
1123088644 M * Bertl /usr/include/linux/capability.h
1123088661 M * Bertl (right above CAP_SYS_RESOURCE)
1123088731 M * Bertl does anything there worry you?
1123088766 M * brc i've read that
1123088768 M * brc but
1123088770 M * brc stuff like
1123088770 M * brc    resources) */
1123088772 M * brc oops
1123088783 M * brc Modify data journaling mode on ext3 filesystem (uses journaling
1123088783 M * brc    resources
1123088787 M * brc is the user able of doing that?
1123088790 M * brc even being inside the vserver ?
1123088798 M * brc there is no /dev/hd* for him
1123088801 M * Bertl most likely not, because of the hopefully missing dev
1123088824 M * Bertl but to be honest I don't know and it is not tested ...
1123088842 M * brc any other way to make the "capped" bind work ?
1123088889 M * Bertl yeah, actually we did work on that some time ago, and IIRC somebody successfully configured it (with a feature patch)
1123088909 M * Bertl but the feature was not tested any further (lack of testers :) so it was removed again ...
1123088918 M * brc i cuold've tested it
1123088918 M * brc hehehe
1123088929 M * Bertl yeah, you could have :)
1123088957 M * Bertl and you can test it in the future if you are interested ...
1123088962 M * brc what about the other stuff on cap_sys_resource ?
1123088963 M * brc yes i can
1123088972 M * brc the server i used to test everything is now on production
1123088977 M * brc but i am gonna get another one soon
1123089019 M * Bertl probably the quota limits and resource limits are the only ones which actually might cause issues
1123089042 M * brc i have no quota
1123089052 M * brc user quota
1123089054 M * brc just vserver quota
1123089068 M * Bertl disk limits are not affected
1123089154 M * brc ok
1123089160 M * brc fine :)
1123089174 M * brc and about the resource limits, would the vserver be able to bypass the limits in /etc/vservers/rlimits ?
1123089201 M * Bertl nope, but the ulimits will be disabled
1123089248 M * brc So i shouldn't worry that much ? :)
1123089251 M * Bertl so for example, setting a core size of 0 will not prevent guest root from changing it back to unlimited
1123089258 M * brc ahh ok
1123089273 M * brc he can change the system ulimits
1123089278 M * brc for the host and all guests
1123089279 M * brc is that it ?
1123089283 M * Bertl and as the core dumping happens from the kernel side, the cores will be able to exceed the disk limits
1123089309 M * Bertl no, the guest can only change the ulimits for guest processes
1123089332 M * brc ahhh ok
1123089380 M * brc so i am gonna keep using it !! hehee
1123089393 M * Bertl yeah, it's probably quite save ... just not perfect
1123089398 M * Bertl *safe
1123089415 M * brc cool
1123089432 M * brc with ngnet vserver would be 100%
1123089460 M * Bertl ... until the next request :)
1123089510 M * brc hehehehe
1123089514 M * brc Ahh.. per vserver user quota :)
1123089590 M * Bertl right :)
1123090961 M * Bertl okay, off again .. back later ...
1123090969 N * Bertl Bertl_oO
1123091603 M * micah I've got a vserver with interface 0 (eth0) configured to have 127.0.0.2, a second one with only a public IP configured,  if I try to communicate with that vserver from the second vserver by using 127.0.0.2 it comes from the public IP of the first, rather than 127.0.0.1
1123092861 J * Aiken ~james@tooax6-140.dialup.optusnet.com.au
1123093007 J * ntrs ntrs@Dardeene-68.188.50.87.charter-stl.com
1123093128 Q * keyser_soze Quit: Abandonando
1123093226 M * brc how are you trying to cmounicate ?
1123093227 M * brc ping ?
1123093539 J * stefani ~stefani@superquan.apl.washington.edu
1123095097 M * micah brc: using the host command to make a DNS query
1123095476 M * micah huh, http://www.karrot-x.net/jamesd/jailVzoneVxen.html doesn't include vservers
1123095533 M * sid3windr yea
1123095537 M * sid3windr he was contacted iirc
1123096435 M * Doener yep, but i didn't get any reply
1123096437 M * Doener :(
1123096552 M * sid3windr although he is on irc!
1123098301 J * kevinp ~kevinp@ny.webpipe.net
1123098391 M * kevinp Does anyone know if you can successfully use the new config style on an old config style vserver?  Is there anything wrong with that?
1123098433 M * kevinp Meaning to take the old vserver filesytem and then slap on the new configs in /etc/vservers for it
1123098518 M * kevinp I have done this, but when I start the vserver, it only mounts /tmp and not /
1123098545 M * kevinp So nothing really starts inside the vserver
1123099296 J * hetche ~lala@dsl-58-6-34-71.nsw.westnet.com.au
1123099692 Q * rhett Ping timeout: 480 seconds
1123100825 M * micah kevinp: i dont think you can do that
1123101502 M * kevinp I tried this http://www.paul.sladen.org/vserver/archives/200411/0164.html and it didn't work either
1123101532 M * kevinp Anything in particular that would make it not work?  It works with the old config
1123101657 M * kevinp Hmm, now the old config doesn't work either
1123101675 M * kevinp I wonder if I screwed it up adding a context quota to it?
1123101717 M * kevinp Is there a way to display what context quotas you currently have enabled?
1123102028 J * yarihm ~yarihm@80-218-5-17.dclient.hispeed.ch
1123102954 J * Matthew-1 ~mlong@grm-nx.grmims.com
1123102990 M * Matthew-1 has anyone gotten vserver to work with suse enterprise 9?
1123103019 M * Matthew-1 i installed it yesterday and i can't seem to get to working
1123103057 M * Matthew-1 i did patch the kernel and that does work,
1123103072 M * Matthew-1 but getting any vserver to run does not
1123103101 M * Matthew-1 anyone have any clues?
1123104989 P * yarihm Leaving
1123105629 N * Bertl_oO Bertl
1123105644 M * Bertl evening folks!
1123105663 M * Bertl Matthew-1: what is/seems to be the problem?
1123105710 M * Bertl kevinp: yes, old style guests work fine with new style configs ...
1123106389 M * Matthew-1 i want to run vserver on a sles9 host
1123106400 M * Matthew-1 i can't seem to get it working
1123106419 M * Bertl so you got a vanilla kernel and patched it with the vserver patches, right?
1123106424 M * Matthew-1 i ran the testme.sh scripted and that returned three errors
1123106429 M * Matthew-1 yes
1123106447 M * Matthew-1 2.6.11
1123106448 M * Bertl could you upload the testme.sh output somewhere (e.g. pastebin.com)?
1123106473 M * Matthew-1 ok
1123106612 M * Matthew-1 i guess i did it
1123106636 M * Matthew-1 i ran ./testme.sh -v
1123106647 M * Matthew-1 and posted it at  pastebin.com
1123106649 M * Bertl hmm, okay, how did you get the tools/patch?
1123106673 M * Bertl I mean, we are at vs2.0-rc9 and 0.30.208 atm ...
1123106687 M * Matthew-1 i followed the instrutions from the kernel 2.6 howto page
1123106692 M * Matthew-1 on the web site
1123106698 M * Bertl and as you are using x86_64, it's probably the best to update
1123106704 M * Matthew-1 ok
1123106711 M * Matthew-1 i'll trie that 
1123106721 M * Matthew-1 try
1123106728 M * Bertl http://vserver.13thfloor.at/Experimental/patch-2.6.12.3-vs2.0-rc9.diff.bz2
1123106751 M * Bertl http://www.13thfloor.at/~ensc/util-vserver/files/alpha/util-vserver-0.30.208.tar.bz2
1123106764 M * Matthew-1 thanks..
1123106776 M * Bertl but you mentioned some guest startup issues
1123106780 M * Matthew-1 i'll let you know if it worked
1123106791 M * Matthew-1 the guest never ran
1123106801 M * Matthew-1 i couldn't get them to even install
1123106803 M * Bertl and although the update is advised, the basic operations should work with your current tools/kernel too
1123106811 M * Matthew-1 Hmmm
1123106814 M * Bertl what did you try?
1123106887 M * Matthew-1 vserver-build  --force -m apt-rpm -n FedoraCore --hostname=grm-test --netdev=eth0 --interface 1.2.3.4 --netmask 255.255.255.0 -- -d suse91
1123106903 M * Matthew-1 i mean fc3
1123106951 M * Bertl hmm, do you have apt-rpm installed?
1123106956 M * Matthew-1 yes
1123106983 M * Bertl hmm, I just saw .. you use vserver-build directly?
1123107004 M * Bertl is this also part of the howto you followed?
1123107011 M * Matthew-1 yes
1123107022 M * Matthew-1 for centos 4
1123107030 M * Bertl I guess I have to review/remove this stuff soon ...
1123107036 M * Matthew-1 ha ha
1123107042 M * Bertl let's try something simple:
1123107062 Q * mef Quit: using sirc version 2.211+KSIRC/1.3.10
1123107073 M * kevinp Matthew-1 which howto did you follow?
1123107110 M * kevinp Bertl/micah, it does work, I had just messed it up somehow by adding a context quota
1123107141 M * kevinp I just rebooted onto rc9 (which works great so far) and I was able to start it without the vdlimit
1123107143 M * Matthew-1 CentOS_HowTo - CentOS 4
1123107157 M * Bertl vserver test42 build -m debootstrap --context 42 --interface 1.2.3.4/24 --hostname=test42.org -- -d sarge
1123107160 M * micah kevinp: ah I misunderstood what you were doing
1123107188 M * Matthew-1 and the Step-by-Step Guide 2.6
1123107206 M * kevinp The CentOS howto does not say to use that line
1123107215 M * Bertl Matthew-1: but I really doubt the step by step guide uses vserver-build directly ...
1123107232 M * Matthew-1 the CentOS_HowTo - CentOS 4 does
1123107236 M * kevinp "vserver min-centos4 build -m yum --hostname domain.com --interface domain=eth0:192.168.0.136/24 --initstyle sysv --context 500 --force -- -d centos4"
1123107306 M * kevinp that is different from what you used
1123107360 M * Bertl Matthew-1: could you give the command sequence I mentioned a try?
1123107419 M * Bertl Matthew-1: maybe with adding  '-m ftp://ftp.at.debian.org/debian/' if required
1123107443 M * Matthew-1 i'm trying to find which one i got it from
1123107505 M * Matthew-1 it seems like it retrieving the debian install
1123107549 M * Matthew-1 oh... when i did the centos 4 one i got this Insufficient server config - no servers found. Aborting.
1123107564 M * Matthew-1 but i think that's a yum problem 
1123107597 M * kevinp Bertl: Is there any reason that vdlimit would work differently on a old vserver with new configs?
1123107603 M * Bertl http://linux-vserver.org/FedoraCore3_HowTo
1123107618 M * Bertl this one uses vserver-build with path directly *urgh*
1123107627 M * Matthew-1 right
1123107650 M * kevinp For some reason when I run the same command that works fine on new vservers it blows away my / and leaves only /tmp
1123107653 M * Bertl well, we'll fix it (the howto)
1123107696 M * Matthew-1 but what about this "Insufficient server config - no servers found. Aborting."
1123107710 M * Matthew-1 that's a yum issue right
1123107716 M * Bertl kevinp: you are talking about vdlimit?! what does that do with your / or /tmp?
1123107735 M * Bertl kevinp: and what command is it you run?
1123107735 M * Matthew-1 i get that when i run "vserver min-centos4 build -m yum --hostname domain.com --interface domain=eth0:192.168.0.136/24 --initstyle sysv --context 500 --force -- -d centos4"
1123107753 M * kevinp well, I run the following command to make my vservers appear to have limited disk drive, hard quota if you will
1123107758 M * kevinp /usr/sbin/vdlimit --xid 300 --set space_total=2000000 --set space_used=`du -s /vservers/testnate/ | awk '{print $1}'` --set inodes_total=2000000 --set inodes_used=`ls -1aRi | awk '/^[0-9]+ / { print $1 }' | sort -u | wc -l` --set reserved=5 /vservers/testnate/
1123107769 M * Bertl Matthew-1: I'd say you have no description/list for the centos guest ...
1123107797 M * kevinp Matthew-1, have you followed all of the steps up to that point in the centos howto?
1123107810 M * kevinp you have to set up a centos distribution etc
1123107818 M * kevinp what OS are you really running on the host?
1123107834 M * Matthew-1 yes.. but i had to find some of the files because i'm running suse
1123107840 M * Matthew-1 sles9
1123107850 M * Bertl kevinp: disk limits (that's the term you are looking for :)
1123107864 M * kevinp Bertl: oh, sorry
1123107886 M * Bertl kevinp: and that does what to your new style/old style guests?
1123107947 M * kevinp on a new vserver, it works perfectly:
1123107952 M * kevinp /dev/hdv1             2.0G  793M  1.1G  43% /
1123107958 M * kevinp none                  150M     0  150M   0% /tmp
1123107987 M * kevinp on a migrated vserver, it works fine (shows the whole drive) until I run the vdlimit command on that context
1123108003 M * kevinp then I just get the /tmp line and no /dev/hdv1 line
1123108014 M * kevinp then it obviously runs out of space
1123108037 M * Matthew-1 how would i run the debian vserver that just installed
1123108042 M * Matthew-1 seems like it's working
1123108052 M * Bertl Matthew-1: vserver test42 start
1123108066 M * Bertl it will probably complain about all kind of hardware stuff
1123108071 M * Bertl but it should start ...
1123108085 M * Matthew-1 No device specified for '/etc/vservers/test42/interfaces/0'
1123108085 M * Matthew-1 Failed to start vserver 'test42'
1123108096 M * Bertl yeah, old tools ...
1123108106 M * Bertl try the following:
1123108117 M * Bertl echo "eth0" >/etc/vservers/test42/interfaces/0/dev
1123108138 M * Matthew-1 ok
1123108156 M * Matthew-1 that seemed to work
1123108162 M * Bertl kevinp: usually if / vanishes, this means that the total is 0
1123108188 M * Bertl Matthew-1: okay, next step is to update the kernel and tools :)
1123108221 M * Bertl kevinp: after executing your command sequence, what does vdlimit show if you query the current settings?
1123108248 M * Matthew-1 ok
1123108250 M * Matthew-1 thanks
1123108346 M * kevinp hmm, I tried both of the internal commands and they give valid results, how to I show the vdlimits?
1123108452 M * kevinp got it
1123108454 M * Bertl internal commands?
1123108460 M * kevinp space_used=0
1123108460 M * kevinp space_total=0
1123108460 M * kevinp inodes_used=0
1123108460 M * kevinp inodes_total=0
1123108460 M * kevinp reserved=100
1123108474 M * Bertl yeah, well, that is not what you want, right?
1123108475 M * kevinp yeah, the commands within the backticks
1123108479 M * kevinp right
1123108503 M * Bertl so maybe you got the path wrong (at the end of the command)? :)
1123108593 M * kevinp ok, so if I run it manually (I had it in a script before) I notice it gives the error "vdlimit: vc_set_dlimit(): Invalid argument"
1123108620 M * Bertl interesting, so something _is_ wrong ...
1123108638 M * Bertl maybe you can prepend an 'echo' to the command sequence?
1123108646 M * kevinp I do not get that error when I run it on a new vserver
1123108667 M * Bertl (and paste the output here)
1123108720 M * kevinp /usr/sbin/vdlimit --xid 300 --set space_total=2000000 --set space_used=3020400 --set inodes_total=2000000 --set inodes_used=3 --set reserved=5 /vservers/testnate/
1123108746 M * Bertl well, space_used > space_total :)
1123108759 M * kevinp oh, man!
1123108767 M * kevinp no wonder!
1123108783 M * Bertl but the inodes_used=3 shows that your script is broken :)
1123108852 M * kevinp hmm
1123108926 M * Bertl something like this: 'find /path/to/vserver -type d -o -links 1 | wc -l' 
1123108939 M * Bertl is probably better but not perfect ...
1123108940 M * kevinp Set it to 4 GB and it works correctly
1123108984 M * kevinp yeah, it's not specifying the path on the ls
1123109064 M * kevinp that error message could be improved then I guess :)
1123109082 M * kevinp Adding the full path fixed the inodes
1123109098 M * Bertl well, invalid argument means you specified something invalid :)
1123109116 M * kevinp lol, I guess you're right!
1123109147 M * Bertl btw, in early versions I accepted any value, just truncated it silently
1123109155 M * Bertl (kernel side)
1123109177 M * kevinp ahh, this way you know you messed something up
1123109180 M * Bertl but that really causes some confusion ...
1123109208 M * kevinp however, wouldn't it be good to not apply the limits at all if they are invalid?
1123109241 M * Bertl yes, and that's probably what is done ...
1123109256 M * kevinp but it changed it to 0
1123109273 M * Bertl give me a second to check the code ...
1123109419 M * Bertl no, it is not set, but the tools first do vc_add_limit()
1123109439 M * Bertl then they call vc_set_limit() which returns with EINVAL
1123109454 M * Bertl leaving the added limit with the zero defaults ...
1123109481 M * Bertl so I'd suggest two changes here ...
1123109504 M * Bertl a) set better defaults in the kernel (i.e. unlimited)
1123109528 M * Bertl b) change the tools to remove the limit if the set fails (and it did the previous add)
1123109632 M * kevinp I would think that b would make more sense, or possibly a third option
1123109659 M * kevinp where it checks for the validity before even doing vc_add_limit()
1123109660 M * Bertl you mean separating add and set?
1123109680 M * Bertl ah, okay :)
1123109703 M * Bertl what tool version do you use, atm?
1123109729 M * kevinp 208
1123109738 M * kevinp without any other patches
1123109752 M * Bertl okay, could you file a bug report on savannah explaining that issue?
1123109758 M * kevinp sure
1123109770 M * Bertl excellent, TIA!
1123111024 M * kevinp bug submitted
1123111110 N * kevinp kevinp|gone